@@ -808,7 +808,6 @@ def _download_and_prepare(self, dl_manager, verify_infos, file_format=None, **pr
808808 # Prepare split will record examples associated to the split
809809 self ._prepare_split (split_generator , file_format = file_format , ** prepare_split_kwargs )
810810 except OSError as e :
811- raise
812811 raise OSError (
813812 "Cannot find data file. "
814813 + (self .manual_download_instructions or "" )
@@ -1239,6 +1238,8 @@ def _prepare_split(self, split_generator, check_duplicate_keys, file_format=None
12391238 generator = self ._generate_examples (** split_generator .gen_kwargs )
12401239
12411240 writer_class = ParquetWriter if file_format == "parquet" else ArrowWriter
1241+
1242+ # TODO: embed the images/audio files inside parquet files.
12421243 with writer_class (
12431244 features = self .info .features ,
12441245 path = fpath ,
@@ -1319,6 +1320,7 @@ def _prepare_split(self, split_generator, file_format=None):
13191320
13201321 generator = self ._generate_tables (** split_generator .gen_kwargs )
13211322 writer_class = ParquetWriter if file_format == "parquet" else ArrowWriter
1323+ # TODO: embed the images/audio files inside parquet files.
13221324 with writer_class (features = self .info .features , path = fpath , storage_options = self ._fs .storage_options ) as writer :
13231325 for key , table in logging .tqdm (
13241326 generator , unit = " tables" , leave = False , disable = (not logging .is_progress_bar_enabled ())
0 commit comments